SPECIAL CONTRIBUTION/NATIONAL HEART ATTACK
ALERT PROGRAM
From the Evidence-based Practice
Center, Division of Clinical Care
Research, New England Medical
Center, Boston, MA.
Dr. Ioannidis is now at the Depart-
ment of Hygiene and Epidemiology,
University of Ioannina School of
Medicine, Ioannina, Greece.
Received for publication July 17, 2000.
Revision received December 29, 2000.
Accepted for publication
February 1, 2001.
This study was conducted by the New
England Medical Center Evidence-
based Practice Center under contract
to the Agency for Healthcare Research
and Quality (formerly, Agency for
Health Care Policy and Research),
contract No. 290-97-0019, Rockville,
MD.
Reprints not available from the
authors.
Address for correspondence: Joseph
Lau, MD, Division of Clinical Care
Research, New England Medical
Center, Box 63, 750 Washington
Street, Boston, MA 02111;
617-636-7670, fax 617-636-8023;
E-mail jlau1@lifespan.org.
Copyright © 2001 by the American
College of Emergency Physicians.
0196-0644/2001/$35.00 + 0
47/1/114904
doi:10.1067/mem.2001.114904
MAY 2001 37:5 ANNALS OF EMERGENCY MEDICINE
Accuracy and Clinical Effect of Out-of-Hospital
Electrocardiography in the Diagnosis of Acute
Cardiac Ischemia: A Meta-Analysis
John P. A. Ioannidis, MD See related articles, p. 450, p. 453, p. 471, and p. 478.
Deeb Salem, MD
Priscilla W. Chew, MPH
Study objective: We sought to evaluate quantitatively the
Joseph Lau, MD
evidence on the diagnostic performance of out-of-hospital ECG
for the diagnosis of acute cardiac ischemia (ACI) and acute
myocardial infarction (AMI) and the clinical effect of out-of-
hospital thrombolysis.
Methods: We conducted a systematic review and meta-analy-
sis of the English-language literature published between 1966
and December 1998 on the diagnostic accuracy of out-of-hospital
ECG and the clinical effect of out-of-hospital thrombolysis. Both
prospective and retrospective studies qualified for the assess-
ment of diagnostic performance. For clinical effect, data from
prospective nonrandomized studies were synthesized sepa-
rately from data from randomized trials. Diagnostic performance
was assessed by using estimates of test sensitivity, specificity,
and diagnostic odds ratios and was summarized by using sum-
mary receiver-operating characteristic curves. Measures of clin-
ical effect included time savings, early ventricular function, early
mortality, and long-term survival.
Results: Diagnostic accuracy was evaluated in 11 studies with
a total of 7,508 patients. Data were available for ACI in 5 stud-
ies and for AMI in 8 studies. For ACI, the random-effects pooled
sensitivity was 76% (95% CI, 54% to 89%), the specificity was
88% (95% CI, 67% to 96%), and the diagnostic odds ratio was
23 (95% CI, 6.3 to 85). The respective figures for AMI were
sensitivity of 68% (95% CI, 59% to 76%), specificity of 97%
(95% CI, 89% to 92%), and diagnostic odds ratio of 104 (95%
CI, 48 to 224). Both in nonrandomized (n=4, total 1,531 patients)
and randomized (n=9, total 6,643 patients) studies, out-of-hos-
pital thrombolysis shortened the time from onset of symptoms
to thrombolytic treatment by 40 to 60 minutes. Data on short-
term ejection fraction were sparse. Hospital mortality was re-
duced by 16% (95% CI, 2% to 27%) among randomized trials,
4 6 1
OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al
and a similar estimate of effect was seen in nonrandomized
studies. There was no clear effect on long-term mortality, but
data were sparse.
Conclusion: Out-of-hospital ECG has excellent diagnostic per-
formance for AMI and very good performance for ACI. Out-of-
hospital thrombolysis achieves time savings and improves
short-term mortality, but the effect on long-term mortality is
unknown.
[Ioannidis JPA, Salem D, Chew PW, Lau J. Accuracy and clinical
effect of out-of-hospital electrocardiography in the diagnosis of
acute cardiac ischemia: a meta-analysis. Ann Emerg Med. May
2001;37:461-470.]
INTRODUCTION
Several studies have addressed the diagnostic performance
of out-of-hospital ECG for evaluating patients with sus-
pected acute myocardial infarction (AMI) and acute car-
diac ischemia (ACI). Furthermore, out-of-hospital ECG
has formed the basis of strategies making use of early
thrombolytic therapy among patients with suspected
AMI before arrival to the hospital. Several studies have
been published since the original National Heart Attack
Alert Program report,1 and the accumulated evidence is
amenable to meta-analysis. The present meta-analysis
therefore addressed 2 major questions. First, what is the
diagnostic accuracy of out-of-hospital ECG? Second,
what is the clinical effect of out-of-hospital initiation of
thrombolysis (in conjunction with out-of-hospital ECG)
compared with hospital initiation of thrombolysis?
Details on methods for the systematic review and gen-
eral study selection criteria for the meta-analysis are pre-
sented in the accompanying synopsis of the evidence re-
port on ACI.2 The characterization of population category
and study quality assessments are also based on algorithms
described in more detail in the same synopsis.2
DIAGNOSTIC PERFORMANCE OF
O U T- O F - H O S P I TA L E L E C T R O C A R D I O G R A P H Y
A total of 11 reports qualified.3-13 There were overlapping
reports for 2 studies. Two reports were derived from the
same population,7,9 and only the first was used. In addition,
the same team also performed a retrospective evaluation
of the diagnostic accuracy of specific ECG changes (not
qualifying for the data synthesis).14 Another 2 reports
stemmed from the Myocardial Infarction Triage and Inter-
4 6 2
vention trial.4,13 The 2 reports provided complementary
data on different diagnostic criteria and outcomes.
Qualifying studies are summarized in Table 1.
All studies targeted populations with chest pain who
required an ECG and did not have any significant exclu-
sion criteria. Nevertheless, in the study by Bertini et al,6
8% of the hospitalized and 24% of the nonhospitalized
patients were excluded because no data were available.
Similarly, in the study by Aufderheide et al7 with 680
enrolled patients, 149 had unsuccessful transmission of
ECG results by telephone, 72 patients were transported to
nonparticipating facilities, and 20 had no medical
records. Finally, in the study by Millar-Craig et al,11 2
phases were planned. During phase I, paramedics were
trained, and the out-of-hospital ECG was not used for
decisionmaking, whereas in phase II, only those para-
medics with accuracy of greater than 80% in their phase I
ECG interpretations participated. In phase II, the out-of-
hospital ECG was used to decide whether the patient
should be directly admitted to the cardiac care unit (CCU).
We excluded 2 studies that also addressed the accuracy
of chest pain protocols by paramedics. One study de-
scribed 25 patients fast tracked by paramedics to the CCU
on the basis of out-of-hospital ECG findings.15 The diag-
nosis of AMI was verified subsequently in 14 patients, but
no information is given on how many patients that were
not fast tracked to the CCU had AMI, and therefore, the
diagnostic characteristics cannot be determined. Another
study addressed the diagnostic accuracy of initiation of a
chest protocol by paramedics, but it was based on a sin-
gle-strip ECG.16 We also excluded 2 studies pertaining to
computerized interpretations of ECGs as part of chest
pain algorithms in the out-of-hospital setting.17,18
Data on diagnostic accuracy for AMI on the basis of the
overall ECG interpretation were available in all 8 qualify-
ing studies, and data on diagnostic accuracy for ACI were
available from 5 studies. Three studies also provided diag-
nostic accuracy data for specific observed ECG changes. To
avoid duplication of patients in the summary calculations,
when more than one set of diagnostic accuracy data ob-
tained with different ECG criteria are reported in a study,
only one set is used, generally the one approaching best
the definition of other studies.
Table 2 summarizes the results of the out-of-hospital
ECG studies considered for diagnostic performance. The
diagnostic odds ratio was 104 (95% CI, 48 to 224) for
AMI and 23 (95% CI, 6.3 to 85) for ACI.
Figure 1 A displays the summary receiver-operating
characteristic curve of 8 studies to assess the diagnostic
performance of out-of-hospital ECG for AMI. The study
ANNALS OF EMERGENCY MEDICINE 37:5 MAY 2001
OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

by Millar-Craig et al11 operated at different sensitivity-
specificity values compared with the other studies be-
cause it used very soft criteria for determining the need
for direct admission to the CCU rather than for diagnosing
AMI. The heterogeneity in the reported sensitivity may be
due to different criteria used to define an abnormal ECG
result.
Figure 1B displays the summary receiver-operating
characteristic results for ACI. The diagnostic accuracy is
very good but far from excellent. Because of different cri-
teria in the definition of coronary ischemia and different
criteria in the definition of an abnormal ECG result, there
was significant heterogeneity of the sensitivity and speci-
ficity estimates.
Three studies addressed the diagnostic accuracy of
specific ECG changes. One study suggested that consid-
eration of reciprocal changes in addition to ST-segment
elevation may be necessary to increase the positive pre-
dictive value to the acceptable range (>90%) for consider-
ation of early out-of-hospital thrombolysis.9 Kudenchuk

Table 1.
Characteristics of studies assessing the diagnostic performance of out-of-hospital electrocardiography.

Subjects Mean Male Inclusion

Evaluated Age Sex (Exclusion) Test
Study (Enrolled) (y) (%) Criteria Criteria Blinding Bias

Aufderheide et al, 151 (166) 64 52 Stable, active CP (SBP <90 mm Hg, AMI: ST-segment elevation ≥0.10 Test and No data
19903 life-threatening arrhythmia) mV in at least 2 of 3 diaphrag- outcome
matic leads, in at least 2 adjacent
precordial leads, or in leads 1 and
aVL. ACI: ST-segment depression,
T-wave inversion, other changes
considered ischemic
Kudenchuk et al, 1,189 (1,189) 57 66 Symptoms <6 h, age <74 y, Para- AMI by computer interpretation on Outcome No data
19914 medic checklist of history and the basis of ST-segment elevation
physical examination criteria
(increased risk for bleeding)
Dalzell et al, 19915 94 (94) 61 67 CP >15 min and <6 h (no data) AMI: ST-segment elevation >0.1 mV Test 3 patients admitted twice
measured 80 ms from J point
Bertini et al, 19916 605 (706) 66 68 No data (no data) No data No data Diagnosis unknown for 95
subjects; 6 subjects died
during intervention
Aufderheide et al, 439 (680) 65 46 Stable, age >18 y (ventricular AMI: ST-segment elevation ≥1 mm No data 265 ECGs excluded (24 poor
19927 arrhythmia, 2o or 3o heart block, in at least 2 anatomically adjacent quality, 149 failed to transmit,
SBP <90 mm Hg) leads 72 subjects transported to
other hospitals, 20 subjects
had no records)
Arntz et al, 19928 1,226 (1,226) No data No data No data (no data) AMI: ST-segment elevation >0.1 mV No data No data
or tall peaked T waves in ≥2
contiguous leads with no BBB or
ventricular pacing
Otto and 428 (439) 65 45 Stable, age >18 y (ventricular AMI: ST-segment elevation ≥1 mm Test and 11 ECGs excluded (8 fully
Aufderheide, 19949 arrhythmia, 2o or 3o heart block, in 2 contiguous leads; other outcome paced, 3 poor quality)
SBP <90 mm Hg) criteria available (see study)
Foster et al, 199410 155 (155) No data No data CP or other referred pain, age AMI: ST-segment elevation ≥1 mm No data No data
>21 y (malignancy, DNR, cardiac in ≥2 contiguous leads and with re-
arrest, unable to give history) ciprocal depression in opposite wall
Millar-Craig et al, 162 (162) 66 80 No data (no data) No data No data No data
199711
Brown, 199712 32 (no No data No data Maryland AMI protocol (no data) No data No data No data
data)
Kudenchuk et al, 3,027 (3,027) 60 66 No data (no data) ACI: ST-segment elevation, ST- Test 10 ECGs excluded (poor quality)
199813 segment depression, T-wave
inversion, Q wave, left BBB
CP, Chest pain; SBP, systolic blood pressure; BBB, bundle-branch block; DNR, do not resuscitate order.

MAY 2001 37:5 ANNALS OF EMERGENCY MEDICINE 4 6 3

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

et al13 evaluated the improved diagnostic performance of
serial out-of-hospital and in-hospital ECGs. Both studies
show the anticipated tradeoff between sensitivity and
specificity when more stringent ECG criteria are used.
Dalzell et al5 provide data on specific ECG changes on a
highly selected population with a very high prevalence of
AMI.
Finally, one study compared computer interpretation
of out-of-hospital ECGs with interpretation by physi-
cians.4 Computer-interpreted ECGs had better specificity
(98% versus 95%) and worse sensitivity (52% versus 66%)
than those undergoing physician interpretation.
The out-of-hospital ECG, as expected, has diagnostic
accuracy for AMI and ACI similar to that of the standard
ECG, which is the gold standard in the management of
patients with chest pain. The accumulated evidence is
substantial, and the data have been gathered from patient
populations with few exclusion criteria. Therefore, an
out-of-hospital ECG should be considered in all patients
with chest pain when first seen by paramedics.
Differences in the stringency of the definition of ECG
abnormalities may result in tradeoffs of sensitivity and
specificity, and the operating point may need to be deter-
mined on the basis of whether these out-of-hospital
ECGs will be used for deciding on the need for out-of-
hospital thrombolysis and what positive predictive value
is therefore needed. Expertise in interpretation may
affect the estimated diagnostic performance. Obviously,
the performance characteristics would be worse if the
ECGs were interpreted by less-experienced practition-

Table 2.
Diagnostic test performance of out-of-hospital ECG studies.

Test Performance
Study Population Prevalence Sensitivity Specificity Study
Study Size Category* of Disease (%) (%) (%) Quality†

Aufderheide et al, 19903 151 I ACI 40 90 53 A

AMI 16 54 99
Kudenchuk et al, 19914 1,189 I ACI 46 — — A
AMI 33 66 95
Dalzell et al, 19915 94 II ACI 92 78 88 B
AMI 51 77 98
Bertini et al, 19916 605 II ACI 52 95 93 B
AMI 26 76 94
Aufderheide et al, 19927‡ 439 I ACI 54 43 87 B
AMI 21 42 99.7
Arntz et al, 19928 1,226 II ACI No data — — B
AMI 34 64 99.5
Otto and Aufderheide, 19949‡ 428 I ACI 60 — — B
AMI 23 60 81
Foster et al, 199410 155 II ACI No data — — B
AMI 14 81 100
Millar-Craig et al, 199711 162 II ACI 66 — — B
AMI 44 99 41
Brown, 199712 32 II ACI No data — — C
AMI 22 100§ 80§
Kudenchuk et al, 199813 3,027 II ACI 53 46II 96II A
AMI 38 — —
Overall 7,508 I ACI 46–92 76 (54–89)¶ 88 (67–96)¶ B
AMI 14–51 68 (59–76)¶ 97 (89–92)¶
*
Category I studies included patients with symptoms suggestive of ACI; category II studies included only patients with chest pain; and category III studies included patients with nondiagnostic ECGs.
See accompanying synopsis article.2
†
Study quality A articles are high-quality, well-documented, unbiased studies. Study quality B articles are fair quality and incompletely documented, with no evidence of significant bias. Study quality
C articles are poor quality and poorly documented, with possibly significant bias. See accompanying synopsis article.2
‡
Aufderheide et al3 and Otto and Aufderheide9 pertain to the same study population.
§
ACI criteria and ST-segment elevation. Other criteria are available (see study).
II
Reporting of the data in the article is inconsistent, incorrect, or both. This study is not used in the quantitative synthesis.
¶
Results of meta-analyses (95% CIs) with random-effects calculations.

4 6 4 ANNALS OF EMERGENCY MEDICINE 37:5 MAY 2001

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

ers, but even experienced staff may miss ECG changes on
the scene.8
C L I N I C A L E F F E C T O F O U T- O F - H O S P I TA L
E L E C T R O C A R D I O G R A P H Y A N D T H R O M B O LY S I S
The clinical effect of out-of-hospital ECG has been
addressed in comparisons of the initiation of thromboly-
sis (on the basis of out-of-hospital ECG) before arrival at
the hospital versus hospital initiation of thrombolysis.
Both prospective nonrandomized studies and random-
ized trials have been performed. On the basis of the avail-
ability of data reported in the literature, we analyzed the
following outcomes: (1) time savings; (2) early differ-
ences in left ventricular function; (3) hospital mortality;
and (4) long-term mortality.
We considered separately randomized trials and non-
randomized studies with prospectively enrolled control
subjects. We excluded observational studies that did not
use control subjects but instead, in most cases, calculated
the observed time delays and speculated on potential time
that could have been saved if thrombolysis was initiated
in the out-of-hospital setting. Such reports19-25 invariably
showed short duration of time to potential out-of-hospital
thrombolysis, as well as possible gains compared with
waiting for arrival to the hospital and obtaining an ECG
there before administering treatment. One more early
study was excluded because although both home- and
hospital-treated patients were considered, their outcomes
could not be separated.26 Finally, the formal meta-analysis
did not consider a study that looked only at the quality of
transmission of out-of-hospital ECGs.27
Retrospective studies were excluded from the evidence
synthesis, but the largest is worth discussing here, given
its sample size.28 This was a retrospective evaluation of
70,763 patients in an AMI registry, including 3,768
patients with out-of-hospital ECGs. Only patients with
AMI presenting to the hospital within less than 12 hours
of the onset of symptoms were included; in-hospital
infarction, transferred-in referrals, and self-transported
patients were excluded. The median time from onset of
symptoms to hospital arrival was prolonged among
patients who had an out-of-hospital ECG (152 versus 91
minutes), although the median time from hospital arrival
to therapy was shortened both for thrombolysis (30 ver-
sus 40 minutes) and for primary angioplasty (92 versus

Figure 1.
A, Summary receiver-operating characteristic curve analysis of out-of-hospital ECG in the diagnosis of AMI. B, Summary receiver-operat-
ing characteristic curve analysis of out-of-hospital ECG in the diagnosis of ACI. Plotted in each of the summary receiver-operating charac-
teristic graphs are individual studies depicted as ellipses. The x- and y-dimensions of the ellipses are proportional to the square root of the
number of patients available to study the specificity and sensitivity, respectively, within the analysis. Also shown is the unweighted summary
receiver-operating characteristic curve limited to the range where data are available. The cross (x) represents the independent random-
effects pooling of sensitivity and specificity values of the studies.

A Sensitivity B Sensitivity
100 100
90 90

80 80

70 70

60 60

50 50

40 40

30 30
20 20

10 10

0 0
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
100–Specificity 100–Specificity

MAY 2001 37:5 ANNALS OF EMERGENCY MEDICINE 4 6 5

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

115 minutes) in the out-of-hospital ECG group. The out- and coronary artery bypass grafting than the control group.
of-hospital ECG group was more likely to have reperfusion In-hospital mortality was 8% in the out-of-hospital group
therapy with thrombolysis or angioplasty, angiography, versus 12% in the control group (P<.001), and the survival

Table 3.
Characteristics of prospective nonrandomized studies of out-of-hospital thrombolysis.

Size, Mean Male

Out-of-Hospital/ Age Sex
Study Design In-Hospital (y) (%) Inclusion Criteria Exclusion Criteria Outcomes

Koren et al, 1985,29 51 55 93

prospective cohort
Weiss et al, 1987,30
case-control study
Roth et al, 1990,31
case-control study
Kereiakes et al, 1990,32
case-control study
Bouten et al, 1992,33
case-control study
Rozenman et al, 1995,34
case-control study
CP, Chest pain; SBP, systolic blood pressure; CABG, coronary artery bypass graft surgery; DBP, diastolic blood pressure; BBB, bundle-branch block; CHF, congestive heart failure.
Age <75 y, CP >30 min and <4 h,
unresponsive to nitrates and
nifedipine, ST-segment elevation
≥0.2 mV, persisting 0.08 s after
J point in ≥2 surface leads and
associated with reciprocal changes
34/84 57 85 CP <4 h, ST-segment elevation
≥2 mm in ≥2 leads and reciprocal
changes, unresponsive to nitrates
and nifedipine
72/44 59 81 Age <73 y, CP >30 min and <4 h,
ST-segment elevation ≥0.1 mV in
≥2 contiguous leads
13/196 No data No data No data
226/220 59 81 Age ≤75 y, CP >30 min and <6 h,
unresponsive to sublingual nitro-
glycerin, ST-segment elevation
≥0.3 mV in ≥2 leads of V1-V6 or
0.2 mV in 2 leads of II/III/aVF with
total ST-segment deviation of ≥1.0
mV in absence of QRS complex of
≥130 ms
114/646 57 82 CP <4 h, ST-segment elevation >2 mm Bleeding tendency, peptic ulcer, or
in ≥2 leads, unresponsive to
sublingual vasodilators
SBP >220 mm Hg, history of severe Mortality, ejection fraction
hypertension, previous CABG,
malignant tumors or terminal
noncardiac illness, contraindi-
cation to streptokinase
Age >76 y, active peptic ulcer, previous Time from symptom onset to
cerebrovascular accident, bleeding treatment, ejection fraction
tendency, SBP >200 mm Hg, history
of AMI
DBP >120 mm Hg, left BBB, history of Time from symptom onset to
CHF or cardiac surgery, terminal treatment, ejection fraction,
illness, bleeding predisposition mortality
No data Time from admission to
treatment
Cardiac massage before evaluation, Time from symptom onset to
stupor, paresis or paralysis, history treatment, mortality
of cerebrovascular accident,
uncontrolled hypertension, bleeding
disorder, fall or head injury after
onset of symptoms
Time from symptom onset to
previous recent cerebrovascular treatment, mortality
accident

Table 4.
Outcomes in nonrandomized studies of out-of-hospital thrombolysis.

Mean (SD) Time to Mean (SD) Mortality

Thrombolysis (min) Ejection Fraction (n/N)*
Study Size Out-of-Hospital Hospital Out-of-Hospital Hospital Out-of-Hospital Hospital Study quality

Roth et al, 199031 118 94 (35) 137 (45) 49 (17) 45 (19) 4/72 3/44 B
Bouten et al, 199233 446 100 (56) 166 (56) No data No data 0/226 6/220 B
Rozenman et al, 199534 760 84 (48) 126 (60) 58 (13) 54 (15) 5/114 11/645 B
Overall 1,324 Mean difference† –50 (49–58)‡ (45–54)‡ Risk ratio† 0.84 (95% CI, 0.17–4.2) B
(95% CI, –32 to –68)
*
Number of subjects who died over total number of subjects.
†
Result of meta-analysis by using random-effects calculations.
‡
Range of studies.

4 6 6 ANNALS OF EMERGENCY MEDICINE 37:5 MAY 2001

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

benefit persisted after adjusting for various predictors of
mortality.
Prospective nonrandomized evidence
Prospective nonrandomized evidence was available
from 4 different teams (Table 3).29-34 Key publications of
the Israeli team that pioneered out-of-hospital thrombol-
ysis include Koren et al,29 Weiss et al,30 Rozenman et al,34
and Weiss et al35 in chronologic order. For data synthesis,
we used the latest update on hard end points.34 Compared
with earlier reports, this report also includes patients
with significant systemic hypertension and patients older
than 75 years. The most recent report does not contain
any data on the prespecified end points.35

Table 5.
Characteristics of randomized trials of out-of-hospital thrombolysis.

Size, Mean Male

Out-of-Hospital/ Age Sex
Study In-Hospital (y) (%) Inclusion Criteria Exclusion Criteria Outcomes

Castaigne et al, 198936
Barbash et al, 199037
Karagounis et al, 199038
Schofer et al, 199039
Kereiakes et al, 199240
McAleer et al, 199241
GREAT, 199242
EMIP, 199343 2,750/2,719
MITI, 199344 and 199645
CP, Chest pain; BBB, bundle-branch block; CHF, congestive heart failure; DBP, diastolic blood pressure; BP, blood pressure; PTCA, percutaneous transluminal coronary arteriography; CAD, coronary
artery disease.
57/43 56 91 Age <75 y, ischemic CP >30 min and
<3 h, unresponsive to nitrates, ST-
segment elevation ≥2 mV in ≥2
standard or 3 precordial leads
43/44 58 80 Age <72 y, CP >30 min and <4 h, ST-
segment elevation of 0.1 mV in ≥2
contiguous leads
34/37 No data No data Age <76 y, CP >20 min, unresponsive
to nitroglycerin, ST-segment
elevation ≥0.1 mV in ≥2 limb leads,
0.2 mV in ≥2 precordial leads, or both
40/38 55 85 Age ≤70 y, CP >30 min and <4 h, ST-
segment elevation ≥2 mm in ≥2
leads for inferior AMI and ≥3 mm in
≥2 precordial leads for anterior AMI
11/11 No data No data No data
43/102 60 78 Symptoms <6 h, ST-segment elevation
≥0.1 mV in ≥2 standard leads, 0.2
mV in at least 2 precordial leads
or both
163/148 63 70 Strong clinical suspicion of AMI,
symptoms >20 min and <4 h, ECG
recorded by general practitioner
61 77 CP >30 min and <6 h, ST-segment
elevation ≥1 mm in 2 limb leads,
elevation >2 mm in ≥2 precordial
leads, or both, history of ACI allowed
175/185 58 82 ≤75 y, CP ≤6 h, ST-segment elevation Risk of serious bleeding including
>1 mm in ≥2 leads, history of CAD
or ACI allowed
History of severe hypertension, Time from symptom onset to
contraindication to thrombolysis treatment, mortality, ejection
fraction
Left BBB, history of CHF, cardiac Time from symptom onset to
surgery, DBP <120 mm Hg, terminal treatment, mortality, ejection
illness, bleeding disposition fraction
Contraindications to thrombolysis Time from hospital arrival to
treatment, time on scene
History of AMI, contraindications to Time from symptom onset to
thrombolysis treatment, mortality, ejection
fraction, wall motion
No data Time from hospital to treatment
Bleeding disorders, anticoagulants, Mortality
active peptic ulcer, recent stroke
or surgery, BP >220/120 mm Hg,
life expectancy <2 y
Thrombolysis <6 mo, surgery or major Mortality
trauma past 10 d, bleeding <6
months, cerebrovascular accident
or neurosurgical procedure <2 months
or known intracranial aneurysm or
neoplasm, history of bleeding disorder,
anticoagulant, risk of pregnancy or
heavy vaginal bleeding, diabetic
proliferative retinopathy, BP >200/
120 mm Hg, recent resuscitation
with chest compression
Anticoagulants, bleeding disorder, Time from symptom onset
recent active peptic ulcer, stroke, to treatment, mortality
surgery or major trauma <6 mo,
recent cardiac massage, BP >200/
120 mm Hg, pregnancy, PTCA <2 wk,
contraindications to thrombolysis
Ejection fraction, infarct size,
history of stroke, recent bleeding mortality, time from symptom
or surgery, known liver or kidney onset to treatment
disease, uncontrolled BP >180/120
mm Hg, hypertension

MAY 2001 37:5 ANNALS OF EMERGENCY MEDICINE 4 6 7

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

The Cincinnati Heart Project addressed the effect of
obtaining an out-of-hospital ECG on the time from reach-
ing the hospital until initiation of thrombolytic therapy.32
Thrombolysis was not given in the out-of-hospital set-
ting. The 13 patients enrolled in the project had average
delays of 36.3 (SD 11.3) minutes versus 62.9 (SD 14.7)
minutes for 196 patients seen in the same facilities where
the out-of-hospital ECG protocol was not applied and
88.8 (SD 54.4) minutes in 211 historic control subjects.
No data are available on time from onset of symptoms to
thrombolysis.
Results on the time from onset of symptoms to treat-
ment are summarized in Table 4. Overall, among the more
than 400 patients given thrombolysis in the out-of-hospi-
tal setting, the time gained was approximately 50 minutes
compared with that of control subjects receiving hospital
thrombolysis.
In 2 studies, ejection fraction in the short term (within
1 week after admission34 or on discharge31) was better in
the out-of-hospital groups but reached statistical signifi-
cance only in the larger study (616 subjects).34 In-hospi-
tal mortality did not differ significantly between the 2
groups, but data were limited, and some heterogeneity
was present (Table 4).
Randomized evidence
A total of 10 published randomized trials pertaining to
the comparison of out-of-hospital with in-hospital strate-
gies were identified.36-45 One small trial was excluded
because the comparison was between thrombolysis in the
CCU and thrombolysis in the ED or at home.46 Seven of
the 9 qualifying trials (Table 5) compared out-of-hospital
with hospital administration of thrombolysis and typi-
cally considered populations of patients with short dura-
tions of symptoms (up to 3-6 hours when stated), age less
than 75 years, and no contraindications to thrombolysis,
as perceived by each trial’s investigators. The other 2 trials
evaluated whether obtaining an out-of-hospital ECG
might increase the time spent in the field38 or might affect
the time to administration of thrombolysis after admis-
sion to the hospital.40 Karagounis et al38 found no in-
crease of in-field time (16.4 versus 16.1 minutes).
Kereiakes et al40 found that the 11 patients randomized to
have an out-of-hospital ECG had a 15-minute reduction

Table 6.
Outcomes in randomized trials of out-of-hospital thrombolysis.

Mean [Median] Ejection

(SD) Time to fraction,
Thrombolysis (min) % (SD) Mortality
Population Out-of- Out-of- Out-of- Study
Study Size Category* Hospital Hospital Hospital Hospital Hospital Hospital Quality†

MITI, 199344 360 IV 92 (58) 120 (49) 53 (12) 54 (12) 10/175 15/185 A
EMIP, 199343 5,469 IV [130] [190] No data No data 266/2,750 303/2,719 A
GREAT, 1992,42 1994,48 311‡ IV [101] [240] No data No data 11/163 17/148 A
and 199749
McAleer et al, 199241 145 IV 138 172 57§ 52§ 1/43 12/102 A
Kereiakes et al, 199240 22 IV No data No data No data No data No data No data A
Schofer et al, 199039 78 IV 85 (51) 137 (50) 51 (10)II 53 (14)II 1/40 2/38 A
Karagounis et al, 199038 71 IV 48 (12) 68 (29) No data No data No data No data A
Barbash et al, 199037 87 IV 96 (36) 132 (42) 48 (15) 48 (15) 1/43 3/43 A
Castaigne et al, 198936 100 IV [131] [180] 57 (no data) 53 (no data) 3/57 2/43 A
Overall 6,643 IV No synthesis possible Range 48–57 Range 48–54 Risk ratio¶ 0.84 A
(95% CI, 0.73–0.98)
*Category I studies included patients with symptoms suggestive of ACI; category II studies included only patients with chest pain; category III studies included patients with nondiagnostic ECGs; and

category IV studies included hospitalized patients only or other inclusion criteria. See accompanying synopsis article.2
†
Study quality A articles are high-quality, well-documented, unbiased studies. Study quality B articles are fair quality and incompletely documented, with no evidence of significant bias. Study quality
C articles are poor quality and poorly documented, with possibly significant bias. See accompanying synopsis article.2
‡
Data from 3 publications of the same study.
§Ejection fraction was measured only in 45% of the patients enrolled.
IIEjection fraction was measured only in 28 patients of each arm.
¶Results of meta-analysis with random-effects calculations.

4 6 8 ANNALS OF EMERGENCY MEDICINE 37:5 MAY 2001

OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al
in the in-hospital time to thrombolytic treatment com-
pared with 9 patients without out-of-hospital ECG.
The time from onset of symptoms to thrombolysis was
mentioned as an outcome in 8 trials (Table 6). Although it
is not possible to arrive at exact summary estimates be-
cause of variability in data reporting, all studies show that
a significant reduction in the time to treatment, ranging
between 20 and 60 minutes, is achieved with out-of-hos-
pital ECG. The largest study found that the difference in
the median time values was 1 hour.43 These results agree
with the data from nonrandomized studies.
Short-term effects on the left ventricular ejection frac-
tion were reported in 5 trials (Table 6).36,37,39,41,44 In
contrast to the results of the nonrandomized studies,
there was no significant difference noted in any of these 5
trials, and a favorable trend was seen only in 2 trials. On
the contrary, usually there were trends for reduced early
mortality (in-hospital or up to 60 days), and the summary
estimate shows a statistically significant 16% reduction in
the risk of death with no heterogeneity among the 7 trials
(Table 6).36,37,39,41-44 The overall estimate of effect is
practically identical to that obtained in the nonrandom-
ized studies.
Long-term effect on mortality was assessed in 4 tri-
als.37,41,42,47 There was substantial heterogeneity. In one
study, 1-year mortality rates were 6.1% and 20%, and 2-
year mortality rates were 9.1% and 30.6% in the out-of-
hospital and hospital arms, respectively. Follow-up pub-
lications on the Grampian Region Early Anistreplase Trial
study also showed a survival benefit with mortality rates
of 10.4% versus 21.6% at 1 year and 25% versus 36% at 5
years.48,49 On the contrary, 2-year survival rates in the
Myocardial Infarction Triage and Intervention study were
89% and 91% for the out-of-hospital and hospital groups,
respectively.45 Three patients in each arm of the Barbash
et al37 trial died over 2 years of follow-up.
The meta-analysis suggests that obtaining an out-of-
hospital ECG does not prolong the in-field time, and if
coupled with administration of thrombolysis in patients
who qualify for this intervention, the intervention may
save approximately three quarters to 1 hour compared
with waiting to give thrombolysis at the hospital. Out-of-
hospital thrombolysis has a modest but significant effect
on early mortality, with approximately 60 patients requir-
ing out-of-hospital treatment compared with in-hospital
thrombolysis to save one additional life in the short term.
Beneficial effects on left ventricular ejection fraction in
the short term have been reported in observational stud-
ies. Obviously ejection fraction data have the disadvan-
tage of selecting out patients who die in the interim. The
MAY 2001 37:5 ANNALS OF EMERGENCY MEDICINE
available data are controversial about survival benefits
from out-of-hospital thrombolysis in the long term.
REFERENCES
1. Selker HP, Zalenski RJ, Antman EM, et al. An evaluation of technologies for identifying
acute cardiac ischemia in the emergency department: a report from the National Heart Attack
Alert Program Working Group. Ann Emerg Med. 1997;29:13-87.
2. Lau J, Ioannidis JPA, Balk EM, et al. Diagnosing acute cardiac ischemia in the emergency
department: a systematic review of the accuracy and clinical effect of current technologies.
Ann Emerg Med. 2001;37:453-460.
3. Aufderheide TP, Hendley GE, Thakur RK, et al. The diagnostic impact of prehospital 12-lead
electrocardiography. Ann Emerg Med. 1990;19:1280-1287.
4. Kudenchuk PJ, Ho MT, Weaver WD, et al. Accuracy of computer-interpreted electrocardiog-
raphy in selecting patients for thrombolytic therapy. MITI Project Investigators. J Am Coll
Cardiol. 1991;17:1486-1491.
5. Dalzell GW, Purvis J, Adgey AA. The initial electrocardiogram in patients seen by a mobile
coronary care unit. Q J Med. 1991;78:227-233.
6. Bertini G, Rostagno C, Taddei T, et al. Evaluation of a mobile coronary care unit protocol in
patients with acute onset chest pain. J Emerg Med. 1991;9(Suppl 1):57-63.
7. Aufderheide TP, Keelan MH, Hendley GE, et al. Milwaukee Prehospital Chest Pain
Project—phase I: feasibility and accuracy of prehospital thrombolytic candidate selection. Am J
Cardiol. 1992;69:991-996.
8. Arntz HR, Stern R, Linderer T, et al. Efficiency of a physician-operated mobile intensive care
unit for prehospital thrombolysis in acute myocardial infarction. Am J Cardiol. 1992;70:417-420.
9. Otto LA, Aufderheide TP. Evaluation of ST segment elevation criteria for the prehospital
electrocardiographic diagnosis of acute myocardial infarction. Ann Emerg Med. 1994;23:17-24.
10. Foster DB, Dufendach JH, Barkdoll CM, et al. Prehospital recognition of AMI using indepen-
dent nurse/paramedic 12-lead ECG evaluation: impact on in-hospital times to thrombolysis in a
rural community hospital. Am J Emerg Med. 1994;12:25-31.
11. Millar-Craig MW, Joy AV, Adamowicz M, et al. Reduction in treatment delay by paramedic
ECG diagnosis of myocardial infarction with direct CCU admission. Heart. 1997;78:456-461.
12. Brown JL Jr. An eight-month evaluation of prehospital 12-lead electrocardiogram monitor-
ing in Baltimore County. Md Med J. 1997;(Suppl):64-66.
13. Kudenchuk PJ, Maynard C, Cobb LA, et al. Utility of the prehospital electrocardiogram in
diagnosing acute coronary syndromes: the Myocardial Infarction Triage and Intervention (MITI)
Project. J Am Coll Cardiol. 1998;32:17-27.
14. Aufderheide TP, Hendley GE, Woo J, et al. A prospective evaluation of prehospital 12-lead
ECG application in chest pain patients. J Electrocardiol. 1992;24(Suppl):8-13.
15. Banerjee S, Rhoden WE. Fast-tracking of myocardial infarction by paramedics. J R Coll
Physicians Lond. 1998;32:36-38.
16. Wuerz RC, Meador SA. Evaluation of a prehospital chest pain protocol. Ann Emerg Med.
1995;26:595-597.
17. Grijseels EW, Deckers JW, Hoes AW, et al. Optimal use of coronary care units: a review.
Prog Cardiovasc Dis. 1995;37:415-421.
18. Aufderheide TP, Rowlandson I, Lawrence SW, et al. Test of the acute cardiac ischemia
time-insensitive predictive instrument (ACI-TIPI) for prehospital use. Ann Emerg Med.
1996;27:193-198.
19. Giovas P, Papadoyannis D, Thomakos D, et al. Transmission of electrocardiograms from a
moving ambulance. J Telemed Telecare. 1998;4(Suppl 1):5-7.
20. Grim PS, Feldman T, Childers RW. Evaluation of patients for the need of thrombolytic ther-
apy in the prehospital setting. Ann Emerg Med. 1989;18:483-488.
21. BEPS Collaborative Group. Prehospital thrombolysis in acute myocardial infarction: the
Belgian eminase prehospital study (BEPS). Eur Heart J. 1991;12:965-967.
22. Aufderheide TP, Haselow WC, Hendley GE, et al. Feasibility of prehospital r-TPA therapy in
chest pain patients. Ann Emerg Med. 1992;21:379-383.
4 6 9
OUT-OF-HOSPITAL ELECTROCARDIOGRAPHY AND ACUTE
CARDIAC ISCHEMIA
Ioannidis et al

23. Weaver WD, Eisenberg MS, Martin JS, et al. Myocardial Infarction Triage and Intervention 45. Brouwer MA, Martin JS, Maynard C, et al. Influence of early prehospital thrombolysis on
Project—phase I: patient characteristics and feasibility of prehospital initiation of thrombolytic mortality and event-free survival (the Myocardial Infarction Triage and Intervention [MITI]
therapy. J Am Coll Cardiol. 1990;15:925-931. Randomized Trial). MITI Project Investigators. Am J Cardiol. 1996;78:497-502.
24. Linderer T, Schroder R, Arntz R, et al. Prehospital thrombolysis: beneficial effects of very 46. McNeill AJ, Cunningham SR, Flannery DJ, et al. A double blind placebo controlled study of
early treatment on infarct size and left ventricular function. J Am Coll Cardiol. 1993;22:1304- early and late administration of recombinant tissue plasminogen activator in acute myocardial
1310. infarction. Br Heart J. 1989;61:316-321.
25. Bossaert LL, Demey HE, Colemont LJ, et al. Prehospital thrombolytic treatment of acute 47. Weaver WD, Cerqueira M, Hallstrom AP, et al. Prehospital-initiated vs hospital-initiated
myocardial infarction with anisoylated plasminogen streptokinase activator complex. Crit Care thrombolytic therapy. The Myocardial Infarction Triage and Intervention Trial. JAMA.
Med. 1988;16:823-830. 1993;270:1211-1216.
26. Fine DG, Weiss AT, Sapoznikov D, et al. Importance of early initiation of intravenous strep- 48. Rawles JM. Halving of mortality at 1 year by domiciliary thrombolysis in the Grampian
tokinase therapy for acute myocardial infarction. Am J Cardiol. 1986;58:411-417. Region Early Anistreplase Trial (GREAT). J Am Coll Cardiol. 1994;23:1-5.
27. Grim P, Feldman T, Martin M, et al. Cellular telephone transmission of 12-lead electrocar- 49. Rawles JM. Quantification of the benefit of earlier thrombolytic therapy: five-year results of
diograms from ambulance to hospital. Am J Cardiol. 1987;60:715-720. the Grampian Region Early Anistreplase Trial (GREAT). J Am Coll Cardiol. 1997;30:1181-1186.
28. Canto JG, Rogers WJ, Bowlby LJ, et al. The prehospital electrocardiogram in acute myocar-
dial infarction: is its full potential being realized? National Registry of Myocardial Infarction 2
Investigators. J Am Coll Cardiol. 1997;29:498-505.
29. Koren G, Weiss AT, Hasin Y, et al. Prevention of myocardial damage in acute myocardial
ischemia by early treatment with intravenous streptokinase. N Engl J Med. 1985;313:1384-1389.
30. Weiss AT, Fine DG, Applebaum D, et al. Prehospital coronary thrombolysis. A new strategy
in acute myocardial infarction. Chest. 1987;92:124-128.
31. Roth A, Barbash GI, Hod H, et al. Should thrombolytic therapy be administered in the
mobile intensive care unit in patients with evolving myocardial infarction? A pilot study. J Am
Coll Cardiol. 1990;15:932-936.
32. Kereiakes DJ, Weaver WD, Anderson JL, et al. Time delays in the diagnosis and treatment
of acute myocardial infarction: a tale of eight cities. Report from the Pre-hospital Study Group
and the Cincinnati Heart Project. Am Heart J. 1990;120:773-780.
33. Bouten MJ, Simoons ML, Hartman JA, et al. Prehospital thrombolysis with alteplase (rt-PA)
in acute myocardial infarction. Eur Heart J. 1992;13:925-931.
34. Rozenman Y, Gotsman MS, Weiss AT, et al. Early intravenous thrombolysis in acute
myocardial infarction: the Jerusalem experience. Int J Cardiol. 1995;49(Suppl):S21-S28.
35. Weiss AT, Leitersdorf I, Gotsman MS, et al. Prevention of congestive heart failure by early,
prehospital thrombolysis in acute myocardial infarction: a long-term follow-up study. Int J
Cardiol. 1998;65(Suppl 1):S43-S48.
36. Castaigne AD, Herve C, Duval-Moulin AM, et al. Prehospital use of APSAC: results of a
placebo-controlled study. Am J Cardiol. 1989;64:30A-42A.
37. Barbash GI, Roth A, Hod H, et al. Improved survival but not left ventricular function with
early and prehospital treatment with tissue plasminogen activator in acute myocardial infarc-
tion. Am J Cardiol. 1990;66:261-266.
38. Karagounis L, Ipsen SK, Jessop MR, et al. Impact of field-transmitted electrocardiography
on time to in-hospital thrombolytic therapy in acute myocardial infarction. Am J Cardiol.
1990;66:786-791.
39. Schofer J, Buttner J, Geng G, et al. Prehospital thrombolysis in acute myocardial infarction.
Am J Cardiol. 1990;66:1429-1433.
40. Kereiakes DJ, Gibler WB, Martin LH, et al. Relative importance of emergency medical sys-
tem transport and the prehospital electrocardiogram on reducing hospital time delay to therapy
for acute myocardial infarction: a preliminary report from the Cincinnati Heart Project. Am Heart
J. 1992;123:835-840.
41. McAleer B, Ruane B, Burke E, et al. Prehospital thrombolysis in a rural community: short
and long-term survival. Cardiovasc Drugs Ther. 1992;6:369-372.
42. Grampian Region Early Anistreplase Trial. Feasibility, safety, and efficacy of domiciliary
thrombolysis by general practitioners: Grampian Region Early Anistreplase Trial. GREAT Group.
BMJ. 1992;305:548-553.
43. European Myocardial Infarction Group. Prehospital thrombolytic therapy in patients with
suspected acute myocardial infarction. The European Myocardial Infarction Project Group. N
Engl J Med. 1993;329:383-389.
44. Weaver WD, Cerqueira M, Hallstrom AP, et al. Prehospital-initiated vs hospital-initiated
thrombolytic therapy. Myocardial Infarction Triage and Intervention Trial. JAMA. 1993;270:1211-
1216.

4 7 0 ANNALS OF EMERGENCY MEDICINE 37:5 MAY 2001